As algorithms continue to have an ever more crucial role in the workplace, lawmakers are concerned over their influence on workers’ rights and wellbeing. A risk-based approach to Artificial Intelligence regulation tries to tackle this challenge, but is it enough?
The pandemic forced many employees to work from home. Along with the work, many also took their employers home, as management tried to supervise what they were up to while on duty, but not in the office. Several software companies now offer products that monitor workers via webcams and track their activities remotely.
Moreover, artificial intelligence has a particular influence on working conditions for workers outside traditional employment, a report by the International Labour Office shows. Platform workers get assigned (or not assigned) work based on automated decision making processes, placing their earnings at the whim of algorithmic decisions.
In April 2021, the EU Commission proposed an AI regulation that aims to prevent the harmful consequences of automated decision-making algorithms. The regulation proposal classifies different applications of artificial intelligence according to the risks they pose.
AI systems used in the workplace to help manage and recruit employees are considered “high-risk” in the proposal. According to the Commission, high-risk AI systems should not be banned, but they should provide “adequate risk assessment and mitigation systems”. Among other measures, they also need to ensure “appropriate human oversight measures to minimise risk.”
Member of the European Parliament Brando Benifei worries that these measures are not enough. If the risk assessment demanded by the regulation can be done in an internal process, he fears that workers will not be protected enough from adverse effects.
“Self-assessment for such invasive practices should not be allowed. Empowering social partners and involving employees is key. We need to make sure that workers do not become an expendable commodity at the service of an algorithm”, Benifei told a EURACTIV event.
Laura Nurski, a research fellow at Bruegel, pointed out that management via algorithms could lead to mental health problems for workers. According to her, algorithms tend to automate decisions about the order of tasks and the speed or rate of work. Algorithms thereby take liberties away from workers that are essential indicators of job quality.
“There is a real risk that these forms of workplace AI could […] set worker wellbeing back 100 years”, Nurski wrote.
In a recent panel discussion, she also pointed out that artificial intelligence could lead to positive developments for workers because boring and repetitive tasks could be replaced. This might bring problems as some workers will lose their jobs. Nurski stressed the importance of providing workers with the right skills for the future of work.
Identifying which employees might need which skills could also be one of the strengths of artificial intelligence applications, said Jens-Henrik Jeppesen of workday, an HR management software provider.
The proposed AI regulation is currently being discussed in European institutions. Both the European Parliament and member states governments will have to agree to the regulation before it can enter into force.
> Watch the full EURACTIV event below on YouTube